Hello Ventuzians!
THE FORUMS ARE CLOSED!

Please join our discord server HERE!! << click me :D

We are shutting our Ventuz Forum, but don't worry, it will all be archived for you to search in if you have a query. From now on, please add all your comments, questions or observations into our Discord Server

Thanks for the great time - see you on discord!!
Dee, Karol, Daniel and the whoooole Product and Support team!

Driver crash

Please report bugs in this group. We will maybe populate a list of known bugs with a possibly modified description at a later time.

Moderator: Support

Post Reply
aster
Posts: 9
Joined: 06 Feb 2018, 12:27

Driver crash

Post by aster » 30 Aug 2018, 23:44

Hello everyone!

I have a cluster of 3 machines (specs listed below) that i am currently testing and evaluating Ventuz, and i am facing a strange issue while trying to sync them up in a cluster. As i am not absolutely sure it is because of Ventuz or some strange/complex reason please feel free to move/remove my post.

Here is the situation:
1: After a fresh boot up of the machines, i go and check for sync signals and so on (nvidia software doesn't seem to remember that they need to sync)
2: On the control machine i am opening Ventuz Configuration Editor, select the group i created and start the default project.
3: After the project start the machines will stutter, Ventuz Presenter will crash along with Nvidia drivers and Device manager will say that the video cards are reporting a problem and cannot start.
4: to fix 3, i disable and then re enable the devices.
5: i go back to 2. Voila ! "Clock from Swapsync" all running all nice and smooth.

To try and fix it I did:
1: Reinstalled the machines numerous times.
2: Tested drivers starting with the latest and going back 7 versions.
3: tested on clean install with QT, DirectX and what not ...
4: different driver profiles and settings...

Unless this is somehow part of the PLE behavior it is .. well strange.

To continue testing and troubleshoot, i installed V5.
This time the result is different. When enabling Swapsync in the configuration V5 starts, but however just display a grid of dots (the ones that then grow in to squares before the default animation starts) But it does not crash neither the driver or the devices. Inspecting the log it says "DirectX Device Failed to enable Framelock"

To continue investigating my friend wrote a small program that uses the swapsync functionality and it is working with no side effects (yes even after a fresh reboot).

Best regards!

Specs:
Ventuz 6.1.1.107 R23109

Machines are as follows:
Dual Intel Xeon E5-2620 V3 Processor (2.40GHz)
16.0GB RAM
1x180GB SSD SATA III
Samsung 960Pro 250GB
Windows 10 Pro 64bit
Nvidia Quadro P4000 8GB + Quadro Sync II
2x NVIDIA NVS 315 1GB Graphics (not in use, they came with the machines so i didn't bother taking them out)

Bellow is the Ventuz Presenter log after the first machine boot and before the above mentioned fix.

First crash after a fresh reboot log:

10:55:41.4 Info : Ventuz Launching Ventuz 6.1 V6.1.1.107 R23109 as 64bit process on Microsoft Windows 8 64bit at 8/30/2018 10:55:41 AM
10:55:41.4 Info : Enviroment Setting CACHE folder to: C:\ProgramData\Ventuz6\Cache
Setting CONFIG folder to: C:\Users\Public\Documents\Ventuz6\Configuration
Setting CONTENT folder to: C:\Users\Public\Documents\Ventuz6\Content
Setting LOG folder to: C:\Users\Public\Documents\Ventuz6\Logs
Setting TEMP folder to: C:\Users\E3-User\AppData\Local\Temp\Ventuz
Setting USER folder to: C:\Users\E3-User\AppData\Roaming\Ventuz6
Setting VPRTEMP folder to: C:\ProgramData\Ventuz6\LocalFS
10:55:41.8 Info : Remote listening to Port 19200 (VPR)
10:55:41.8 Info : CLI listening to Port 19201
10:55:41.8 Info : OscRemote listening to Port 19230
10:55:41.8 Info : Remote Remoting2 listening to Port 19200
10:55:41.8 Info : Remote Remote Render-To-Disk listening to Port 19200
10:55:41.8 Info : Input Subsystem Using single machine setup.
10:55:41.8 Info : Input Subsystem Windows Touch is active on TouchGroup Default
10:55:41.8 Info : Input Subsystem TUIO listening to Port 3333; active on TouchGroup Default
10:55:41.8 Info : Input Subsystem Generating Touches for Mouse on TouchGroups Default (LMB) and None (RMB)
10:55:41.8 Info : Device Selecting Device #0)
10:55:41.8 Info : Runlevels enter runlevel Device
10:55:41.8 Info : graphics setting displaymode 1920 x 1080 at 60000 / 1000 Hz
10:55:41.9 Info : Runlevels enter runlevel Resource
10:55:41.9 Info : graphics Rendertarget format Null not available, replaced with RGBA_UN8
10:55:42.0 Info : graphics warming up 34 shaders
10:55:42.0 Info : Runlevels enter runlevel ScreenSize
10:55:42.2 Info : Ventuz Rendering on a NVidia GPU.
10:55:42.2 Debug : RenderSetup Machine Production RenderTarget Size = 1920 x 1080
10:55:42.2 Info : Clock Local timing 60.00fps
10:55:42.4 Info : ClusterActionManager Initialized on Port 19333
10:55:42.4 Info : ClusterActionManager Cluster Machine IPs updated to: 192.168.1.202, 192.168.1.210, 192.168.1.209,
10:55:49.1 Fatal : Ventuz Unhandled Exception occurred AltonaLite.Assertion : e:\jenkins_workspace\ventuz 6.01 maintenance\ventuz.pure\engine\device.cpp(9091): Assertion
10:55:49.1 Error : graphics failed create directx device (SwapChain->GetBuffer), error 887a0001 (1)

After a fix:

11:05:49.3 Info : Ventuz Launching Ventuz 6.1 V6.1.1.107 R23109 as 64bit process on Microsoft Windows 8 64bit at 8/30/2018 11:05:49 AM
11:05:49.3 Info : Enviroment Setting CACHE folder to: C:\ProgramData\Ventuz6\Cache
Setting CONFIG folder to: C:\Users\Public\Documents\Ventuz6\Configuration
Setting CONTENT folder to: C:\Users\Public\Documents\Ventuz6\Content
Setting LOG folder to: C:\Users\Public\Documents\Ventuz6\Logs
Setting TEMP folder to: C:\Users\E3-User\AppData\Local\Temp\Ventuz
Setting USER folder to: C:\Users\E3-User\AppData\Roaming\Ventuz6
Setting VPRTEMP folder to: C:\ProgramData\Ventuz6\LocalFS
11:05:49.6 Info : Remote listening to Port 19200 (VPR)
11:05:49.6 Info : CLI listening to Port 19201
11:05:49.6 Info : OscRemote listening to Port 19230
11:05:49.6 Info : Remote Remoting2 listening to Port 19200
11:05:49.6 Info : Remote Remote Render-To-Disk listening to Port 19200
11:05:49.6 Info : Input Subsystem Using single machine setup.
11:05:49.6 Info : Input Subsystem Windows Touch is active on TouchGroup Default
11:05:49.6 Info : Input Subsystem TUIO listening to Port 3333; active on TouchGroup Default
11:05:49.6 Info : Input Subsystem Generating Touches for Mouse on TouchGroups Default (LMB) and None (RMB)
11:05:49.6 Info : Device Selecting Device #0)
11:05:49.6 Info : Runlevels enter runlevel Device
11:05:49.6 Info : graphics setting displaymode 1920 x 1080 at 60000 / 1000 Hz
11:05:49.6 Info : graphics get-focus-hack activated
11:05:49.7 Info : Runlevels enter runlevel Resource
11:05:49.7 Info : graphics Rendertarget format Null not available, replaced with RGBA_UN8
11:05:49.7 Info : graphics warming up 34 shaders
11:05:49.8 Info : Runlevels enter runlevel ScreenSize
11:05:49.9 Info : Ventuz Rendering on a NVidia GPU.
11:05:49.9 Debug : RenderSetup Machine Production RenderTarget Size = 1920 x 1080
11:05:49.9 Info : Clock Local timing 60.00fps
11:05:50.1 Info : ClusterActionManager Initialized on Port 19333
11:05:50.1 Info : ClusterActionManager Cluster Machine IPs updated to: 192.168.1.202, 192.168.1.210, 192.168.1.209,
11:05:59.3 Info : videoengine Video Engine Streams
11:05:59.3 Info : videoengine DirectX 0:2 DirectX output(NVIDIA Quadro P4000 : DISPLAY1) : status Running
11:05:59.3 Info : videoengine Audio Crosspoints

aster
Posts: 9
Joined: 06 Feb 2018, 12:27

Re: Driver crash

Post by aster » 04 Sep 2018, 09:34

OK.
Seems i am not the only one with no clue whats happening....

User avatar
Eric_RD
Posts: 103
Joined: 04 Jun 2014, 14:01
Contact:

Re: Driver crash

Post by Eric_RD » 04 Sep 2018, 21:52

Why do you enable swap sync?
If the machines are Framelocked (view System topology in nvidia driver or the LEDs on the Sync board (should be green))
and the Machines are in one Ventuz Group (config editor)
and you have a cluster license (or PLE)
the machines are perfectly in Sync.
The lower Ventuz ID will be timing master (lets Say ID1) the other IDs will receive the Ventuz framecouter which you can see with live options -> performance statistics->the timing should be green on "timing slaves" showing something like "frame 837236 from 01.01"

Hope that helps... if you don't lose warranty mabe you can take out the nvs cards if you don't need them anyway

Post Reply